Atos Previews Energy-Efficient, AI-Augmented Hybrid Supercomputer

The BullSequana XH3000 — designed for both traditional digital simulation and accelerated hybrid computing — embraces NVIDIA CPU, GPU and networking technologies to deliver a new level of performance for HPC, AI and quantum applications.
by Dion Harris

Stepping deeper into the era of exascale AI, Atos gave the first look at its next-generation high-performance computer.

The BullSequana XH3000 combines Atos’ patented fourth-generation liquid-cooled HPC design with NVIDIA technologies to deliver both more performance and energy efficiency.

Giving users a choice of Arm or x86 computing architectures, it will come in versions using NVIDIA Grace, Intel, AMD or SiPearl processors. For accelerated computing, it supports nodes with four NVIDIA Tensor Core GPUs.

The XH3000 also flexibly employs network options including NVIDIA Quantum-2 InfiniBand and NVIDIA ConnectX-7 InfiniBand and Ethernet adapters to scale these powerful computing nodes to HPC systems capable of 10 mixed precision AI exaflops.

Hybrid HPC+AI+Quantum System

The result is a flexible, hybrid computing platform capable of running the most demanding HPC simulations, AI jobs and even emerging workloads in quantum computing.

The BullSequana XH3000 “will no doubt enable, through the gateway of exascale, some of the key scientific and industrial innovation breakthroughs of the future,” said Rodolphe Belmer, CEO of Atos, in a virtual event revealing the system.

With customers in more than 70 countries, Atos is #1 in supercomputing in Europe, India and South America and especially renowned in France, where it maintains its headquarters as well as a manufacturing and R&D base.

A Broad Collaboration

Optimizations for the XH3000 were among the first projects for EXAIL, the joint Excellence AI Lab that Atos and NVIDIA announced in November.

John Josephakis, global vice president of sales and business development for HPC/supercomputing at NVIDIA, congratulated the team behind the system in a video message.

“By combining the well-known expertise Atos has with NVIDIA AI and HPC technologies and work at our joint lab, this platform will allow researchers to get significant insights much faster to grand challenges both in supercomputing and industrial HPC,” he said.

EXAIL’s work spans climate research, healthcare and genomics, quantum computing, edge AI/computer vision and cybersecurity. Its researchers can access application frameworks such as NVIDIA Clara for healthcare and NVIDIA Morpheus for security as well as the NVIDIA cuQuantum SDK for quantum computing and the NVIDIA HPC SDK that runs hundreds of scientific and technical applications.

A Long, Productive Relationship

Atos built one of Europe’s first supercomputers to employ the NVIDIA Ampere architecture, the JUWELS Booster at the Jülich Supercomputing Center. It uses 3,744 NVIDIA A100 Tensor Core GPUs to deliver 2.5 exaflops of mixed-precision AI performance.

To provide a deeper understanding of climate change, Atos and NVIDIA researchers will run AI models on the system, currently ranked No. 8 on the TOP500 list of the world’s fastest supercomputers. Julich researchers used the system in April to conduct a state-of-the-art quantum circuit simulation.

Last year, Atos led deployment of BerzeLiUs, a system built on the NVIDIA DGX SuperPOD and Sweden’s largest supercomputer. The company also has delivered supercomputing infrastructure in Europe, India and South America based on NVIDIA DGX systems.

Next up, Atos is building Leonardo, a supercomputer at the Italian inter-university consortium CINECA. It will pack 14,000 NVIDIA A100 GPUs on an NVIDIA Quantum InfiniBand network and is expected to become the world’s fastest AI supercomputer, capable of 10 exaflops of mixed-precision AI performance.

With the first glimpse of the BullSequana XH3000, it’s clear there’s much more to come from the collaboration of Atos and NVIDIA.